MINERALOGY and U/Pb PHOSPHATE DATING of the LOS ANGELES MARTIAN DIABASE

50th Lunar and Planetary Science Conference 2019 (LPI Contrib. No. 2132) 2952.pdf

MINERALOGY AND U/Pb PHOSPHATE DATING OF THE LOS ANGELES MARTIAN DIABASE. B. J. Wilson1, C. R. M. McFarlane1, and J. G. Spray2, 1Earth Sciences, 2Planetary and Space Science Centre, University of New Brunswick, Fredericton, New Brunswick, E3B 5A3, Canada ([email protected])

Introduction: The Los Angeles (LA) Martian me- They are dominated by ferromerrillite with minor apa- teorite was discovered in 1999 by Robert Verish in a tite. In the crystallization sequence, merrillite typically pre-existing rock collection. LA is a slightly to negligi- precedes apatite [2], but in some instances, merrillite bly weathered medium-grained shergottite consisting of and apatite crystallized simultaneously. two stones: LA001 (453 g) and LA002 (245 g) [1]. Results: A Micro-XRF was used to map phos- They primarily comprise pyroxene and maskelynitized phates within 5 available thick sections (LA001–T1 plagioclase (>80%), with phosphates contributing through LA001–T5) pertaining to stone 1. Data collec- ~2.5% in both stones [2]. These phosphates contain tion focused on the most phosphate-rich thick sections: uranium, allowing in situ Laser Ablation Inductively LA001–T2 and LA001–T5 (Figure 1), except for the Coupled Plasma Mass Spectrometry (LA-ICP-MS) to Raman. The Raman was deployed on a separate Los be performed for the first time on LA. Three spectros- Angeles polished thin section (loaned from the Smith- copy techniques were deployed to characterize the sonian Institute: sample 7059-1) due to limited optical phosphates: A Field Emission Scanning Electron Mi- resolution of the Raman microscope in reflected light croscope (FESEM) equipped with an Energy Disper- when using the thick sections. sive Spectrometer (EDS), a Raman Spectrometer, and an Electron Microprobe equipped with a Wavelength Dispersive Spectrometer (WDS). LA-ICP-MS was deployed to determine the phosphate age, and to ap- praise the ability of the LA phosphates to resist shock resetting due to impact, and ejection at ~3 Ma [3]. Mineralogy: Modal analysis indicates that Los Angeles is composed chiefly of maskelynitized plagioclase and pyroxene, totalling ~89 wt% for stone 1, and ~81 wt% for stone 2. Los Angeles has significantly more maskelynite (~45%) than many other Martian shergottites, such as Shergotty (~23%) and Zagami (~22%) [1]. Modal abundances for other minor phases exhibited are silica (~3-4%), ulvospinel (~3%), fayalite (~3%) K-rich feldspathic glass (~2%), ferromerrillite (~1.5%), apatite (~1%), pyrrhotite (~0.6%), and ilmen- ite (~0.3%) [2]. About 5-10% of LA displays a symplectitic texture of ~50-200 m-sized vermicular intergrowths of fayalite, ferroan clinopyroxene (heden- bergite), and silica in a 2:2:1 ratio, termed “pyroxferroite breakdown material” (PBM) [1]. The conversion of pyroxferroite to PBM indicates that LA cooled from igneous temperatures (~1100-900˚C) in a period ex- ceeding 3 days, otherwise metastable pyroxferroite would be expected [1]. Stone 1 and stone 2 exhibit minor differences. Stone 2 contains more late-stage

melt products (fayalite, silica, and K-rich feldspathic glass) and is mainly subophitic [1]. Stone 1 texturally Figure 1: Micro-XRF images for thick sections more closely resembles a plagioclase-pyroxene ortho- LA001–T5 (top) and LA001–T2 (bottom). Phosphates cumulate [1]. It was inferred that 95-97% of LA had are indicated by the green/turquoise colour for both crystallized before phosphates crystallization, which is samples. Red areas are Ti-rich. borne out by textural relations in our samples [4]. The phosphates occur as relatively large crystals FESEM and EDS: 149 analyses were obtained from (>>100m), which are amenable to laser ablation. 37 phosphate crystals via EDS. Based on totals from 50th Lunar and Planetary Science Conference 2019 (LPI Contrib. No. 2132) 2952.pdf

99 to 101 wt%, 91 analyses were approved; 78 of LA-ICP-MS: Utilizing the U/Pb system, LA-ICP- which were ferromerrillite (Ca9NaFe(PO4)7); 13 were MS analysis was performed to obtain an age for the chlor- or fluorapatites. Most existing literature cites phosphates, and to test their susceptibility to U-Pb re- merrillite Ca9NaMg(PO4)7 or whitlockite setting during ejection-related shock. Large (~80 m) Ca9(MgFe)(PO4)6PO3OH as the major phosphate phas- craters were utilized by the LA-ICP-MS and were es in Martian meteorites [e.g., 1, 2]. However, due to placed to avoid surface fractures as best as possible. enriched Na2O (~1 wt%) and FeO (~5 wt%), combined The collected points were plotted on an inverse Con- with depleted MgO (0.8 wt%), ferromerrillite is con- cordia plot with a data-point error ellipse of 2. Utiliz- firmed here. Ferromerrillite is a trigonal, R3c space ing all 117 data points collected, an age of 171  31 group mineral, and is the dominant phosphate phase in Ma was produced. Additionally, to test the shock re- LA [5]. sistance of the phosphate phases, phosphates within, or Raman Spectroscopy: Due to the inability of EDS neighboring shock melt pockets were separately target- and WDS techniques to detect hydrogen, Raman spec- ed to evaluate their age. From this, an age of 147  42 trometry was deployed to further distinguish ferromer- Ma was produced, providing evidence that they are rillite from whitlockite. Using 50 acquisitions on an robust, and shock resetting has not occurred (i.e., due assumed ferromerrilllite crystal, Figure 2 was pro- to ejection at ~3 Ma). After removing outliers, and duced, ruling out whitlockite. The apatites were also including phosphate points within the shock melt pock- investigated for hydroxylapatite via Raman, but the ets, a refined age of 170  16 is produced (Figure 2), hydrogen peak was not present in the signature. This showing good agreement with previously reported Rb- shows that both phosphate phases are anhydrous. Sr and Sm-Nd ages [3]. This confirms an igneous crystallization age of ~170 Ma for Los Angeles via in situ LA-ICP-MS.

Figure 2: Raman spectrum for ferromerrillite (top) compared to merrillite from the Ruff Database. Ac- quired data comprise 50 acquisitions with baseline sub- tracted. No hydrogen peak is apparent.

Figure 3: After excluding outliers, an inverse Concor- Microprobe and WDS: As ferromerrillite has not dia plot was utilized to produce a refined age of 170  been extensively studied, the microprobe was deployed 16 Ma. to obtain higher precision analyses of the phosphates. A total of 110 analyses were obtained; 67 ferromerril- References: [1] Rubin A. E. et al. (2000) Geology, lite and 24 apatite analyses, with 24 points rejected for 28, 1011-1014. [2] Warren P. H. et al. (2004) Meteor- existing outside the 98-102% accepted totals range. Ba, ics. Planet. Sci., 39, 137-156. [3] Nyquist L. et al. Sr, and Al were found to be trace constituents in both (2001) Space Science Reviews, 96, 105-164. [4] phases of phosphates. In comparison to the EDS re- Xirouchakis D. M. et al. (2002) Geochim. Cosmochim. sults, the largest change in the WDS data corresponds Acta, 66, 1867-1880. [5] Britvik et al. (2016) Europ. J. to the average decrease in P2O5 and average increase in Mineral. 28, 125-136. CaO and MnO. In LA001-T5, 16 phosphate points (9 ferromerrillite and 7 apatite) were tested for LREEs, U and S. Trace amounts of S existed in all 7 apatites; the rest were below detection limit for both phosphates.